app.py

# app.py - Complete Flask Backend
from flask import Flask, request, jsonify, send_from_directory
import google.generativeai as genai
from dotenv import load_dotenv
import os
from flask_cors import CORS
import markdown2
import re
from gtts import gTTS
import uuid
import logging

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Load environment variables
load_dotenv()

# Configuration
AUDIO_FOLDER = os.path.join('static', 'audio')
os.makedirs(AUDIO_FOLDER, exist_ok=True)
MAX_AUDIO_LENGTH = 5000  # characters

# Initialize Flask app
app = Flask(__name__, static_folder='static')
CORS(app)

# Enhanced Gemini System Instruction
SYSTEM_INSTRUCTION = """
You are AstroChat, an advanced AI assistant with voice capabilities. Follow these guidelines:

1. Voice Responses:
- When users request audio (e.g., "read this", "speak aloud", "audio version"), include [AUDIO] in response
- Structure responses for optimal TTS:
  * Short sentences (12-15 words)
  * Pause between paragraphs
  * Spell out complex terms

2. Content Formatting:
- Code: Explain → Format in markdown
- Lists: Use bullet points
- Quotes: Provide attribution
- Math/Science: Explain symbols verbally

3. Interaction Style:
- Friendly but professional
- Ask clarifying questions
- Admit knowledge limits
- Offer follow-up suggestions

4. Special Cases:
- Acronyms: Spell out first use
- Names: Provide pronunciation hints
- Technical terms: Give simple definitions
"""

# Initialize Gemini
genai.configure(api_key=os.getenv("GEMINI_API_KEY"))
model = genai.GenerativeModel(
    'gemini-1.5-flash',
    system_instruction=SYSTEM_INSTRUCTION
)

def process_response(text):
    """Process AI response for audio triggers and markdown conversion"""
    audio_requested = '[AUDIO]' in text
    clean_text = text.replace('[AUDIO]', '').strip()
    
    # Convert markdown to HTML with enhanced processing
    extras = [
        "fenced-code-blocks", 
        "tables",
        "code-friendly",
        "cuddled-lists"
    ]
    html = markdown2.markdown(clean_text, extras=extras)
    
    # Enhanced code block styling
    html = re.sub(
        r'<pre><code(.*?)>', 
        r'<pre class="code-block"><code\1>', 
        html
    )
    
    # Improve link handling
    html = re.sub(
        r'<a href="(.*?)">(.*?)</a>',
        r'<a href="\1" target="_blank" rel="noopener">\2</a>',
        html
    )
    
    return {
        "response_html": html,
        "response_text": clean_text,
        "audio_requested": audio_requested
    }

@app.route('/chat', methods=['POST'])
def handle_chat():
    try:
        data = request.json
        user_message = data.get('message', '').strip()
        
        if not user_message:
            return jsonify({"error": "Empty message"}), 400
        
        # Detect audio requests
        audio_triggers = [
            "read aloud", "speak this", "audio please",
            "say it", "voice response", "read this",
            "can you speak", "tell me aloud"
        ]
        explicit_audio = any(
            trigger in user_message.lower() 
            for trigger in audio_triggers
        )
        
        # Generate response
        response = model.generate_content(user_message)
        processed = process_response(response.text)
        
        # Force audio if explicitly requested
        if explicit_audio:
            processed["audio_requested"] = True
        
        return jsonify(processed)
        
    except Exception as e:
        logger.error(f"Chat error: {str(e)}")
        return jsonify({
            "error": "I encountered an error",
            "details": str(e)
        }), 500

@app.route('/generate-audio', methods=['POST'])
def handle_audio():
    try:
        data = request.json
        text = data.get('text', '').strip()
        
        if not text:
            return jsonify({"error": "No text provided"}), 400
        
        # Enhanced text cleaning
        clean_text = re.sub(r'[\*_`#\[\]]', '', text)  # Remove markdown
        clean_text = re.sub(r'\s+', ' ', clean_text).strip()
        
        # Safe truncation
        if len(clean_text) > MAX_AUDIO_LENGTH:
            clean_text = clean_text[:MAX_AUDIO_LENGTH]
            clean_text += "... [content truncated]"
        
        # Generate unique filename
        filename = f"audio_{uuid.uuid4()}.mp3"
        filepath = os.path.join(AUDIO_FOLDER, filename)
        
        # Generate speech with enhanced parameters
        tts = gTTS(
            text=clean_text,
            lang='en',
            slow=False,
            lang_check=False,
            pre_processor_funcs=[
                lambda x: re.sub(r'([a-z])([A-Z])', r'\1 \2', x)  # Handle camelCase
            ]
        )
        tts.save(filepath)
        
        return jsonify({
            "audio_url": f"/audio/{filename}",
            "text_length": len(clean_text)
        })
        
    except Exception as e:
        logger.error(f"Audio error: {str(e)}")
        return jsonify({
            "error": "Audio generation failed",
            "details": str(e)
        }), 500

@app.route('/audio/<filename>')
def serve_audio(filename):
    try:
        return send_from_directory(AUDIO_FOLDER, filename)
    except FileNotFoundError:
        return jsonify({"error": "Audio file not found"}), 404

@app.route('/')
def serve_index():
    return send_from_directory('static', 'index.html')

@app.route('/<path:path>')
def serve_static(path):
    return send_from_directory('static', path)

if __name__ == '__main__':
    port = int(os.environ.get('PORT', 7860))
    app.run(host="0.0.0.0", port=port)